PTQ Q1 2024 Issue

Sizing preflash devices The ideal preflash drum is horizontal because the cross-sectional area for flashed crude flow is larger. However, the plot area available may be insufficient for a horizontal vessel, and a vertical vessel would be preferred, particularly in revamps. Because the cross-sectional area for flashed crude flow in the vertical drum is smaller, it will require more vessel height to retain the foam. A preflash drum, as well as the bottom section of a preflash tower, must have sufficient height to contain the foam to prevent the rising vapour from carrying foam (flashed crude) into the drum overhead or the upper preflash column trays. Conventional trays will not break foams and will flood upon foamovers. Barber and Wijn⁸ presented the results of experiments on pilot and full-scale preflash drums pro- cessing various crudes at several operating conditions. Based on these data, Barber and Wijn derived a model and correlation that can be used to design crude preflash- ing devices. According to their tests, the key sizing criterion for the drum or the tower bottom section is the flashed crude downward superficial velocity, which should be low enough to allow the foam to be retained inside the drum or in the bottom of the tower. As long as there is suffi- cient height above the maximum clear liquid level (HHLL), foamover is unlikely. The smaller the cross-sectional area for flashed crude, the higher the foam level inside. If the disengaging height above the maximum clear liquid level is small, then the superficial velocity must be low. Conversely, if the available height to retain the foam is high, then a higher superficial velocity can be tolerated before the foam is no longer retained. The Figure 7 correlation8 applies to a crude oil superficial liquid velocity in the range of 15 to 30 gpm/ft2. We have had excellent experience with this correlation, and it seems to extrapolate well to values up to 60 gpm/ft2, but extrapolation beyond 30 gpm/ft2 cannot be done with confidence. It is important to check the preflash tower or drum size for the lightest (and heaviest) crudes intended to run, including start-up conditions with slop reprocessing.3 Some literature sources advocate a residence time crite- rion for preflash drum sizing (for example, three minutes). The authors’ experience strongly favours the downward velocity criterion above but does not support a residence time criterion. In an attempt to alleviate severe foaming in one preflash tower (not a Fluor design), a refinery added a horizontal drum in parallel with the tower’s vertical sump. The liquid and vapour spaces of the drums were con- nected by large lines, as shown in Figure 8 . Doubling the residence time with the addition of the horizontal drum did very little to alleviate the foaming. The downward velocity remained unchanged as all the crude was fed to the flash zone of the tower. Drum or tower inlet devices Barber and Wijn⁸ also found that foaming in the preflash drum or tower is strongly affected by the inlet device. The inlet device needs to be adequately designed and carefully checked.

6

5

4

3

2

1

0

0

5

10

15

20

25

30

35

Supercial liquid velocity (gpm/ft)

Figure 7 Foam height vs superficial liquid velocity Basis Figure 5, Reference 8

In preflash towers, even with vortex foam separators, some entrainment from the flash zone remains, which would normally be washed down by the reflux. If the reflux on the tray above runs dry or very low, this entrainment may lead to a dark product, as we have experienced. Lastly, good instrumentation is invaluable in alleviating and managing foamovers. Pressure, levels, flow rates, and, most of all, temperatures can provide operators with an early indication of the onset of foamovers. Instability in drum level, pressure, and flashed crude flow rate forewarns the approach of a foaming episode. Reliable measurement of the drum level has been inherently problematic due to foam formation. Seeing a ‘split level’ in the sight glass indicates likely foam.3 A nuclear-level device has been advocated for monitoring the preflash drum or tower level.7 Watching the preflash drum overhead temperature and the bottom pump for cavitation is also invaluable for detecting foamovers. With preflash towers that have a kero side draw, it is imper- ative to closely monitor the colour of the draw because that is where the foamover is first seen.⁷

Naphtha

Reux

Crude feed

Preash crude

Figure 8 Adding this horizontal drum did not alleviate foaming

86

PTQ Q1 2024

www.digitalrefining.com

Powered by